基于介电特性的生鲜牛奶含水率检测方法

doi:10.6048/j.issn.1001-4330.2025.04.022

摘要/Abstract

摘要：

【目的】探索牛奶介电特性和含水率之间的关系,为有效预测生鲜牛奶含水率、实现含水率快速检测提供参考。【方法】采用矢量网络分析仪和同轴探头测量2~20 GHz频率范围内牛奶样品的介电常数(ε')和介质损耗因数(ε″)。采用偏最小二乘回归(PLSR)、支持向量回归(SVR)和基于粒子群优化的最小二乘支持向量回归(PSO-LSSVR)3种建模方法,依次以ε'和ε″为变量建立6种数学模型对牛奶含水率进行预测并选优,精准预测牛奶含水率。【结果】随着频率的增加,ε'呈逐渐减小的趋势,ε″呈逐渐增大的趋势。6种模型中基于PSO-LSSVR方法下以ε″为变量建立的模型具有最好的含水率预测性能,其R²和RMSE分别为0.996 3和0.001 3。【结论】在2~20 GHz,随着频率的增加,ε'呈逐渐减小的趋势,而ε″则逐渐增加,介电特性可有效地预测牛奶的含水率。

关键词: 牛奶; 含水率; 介电特性; 粒子群算法; 支持向量回归; 偏最小二乘回归

Abstract:

【Objective】 Explore the relationship between milk dielectric characteristics and moisture content,In order to scientifically and effectively predict the moisture content of raw milk, realize rapid detection of moisture content. 【Methods】 The dielectric constant (ε') and dielectric loss factor (ε″) of milk samples between 2 and 20 GHz were measured using vector network analyzer and coaxial probe. PLSR, SVR and PSO-LSVR were used to use ε'or ε″ as variables to predict the milk from 86.9% to 92.9% and select the best, so as to achieve the accurate prediction of milk moisture content. 【Results】 The results showed that as the frequency increased, ε' gradually decreased, while ε'' gradually increased. Among the six models, ε″ based on the PSO-LSSVR method had the best water content prediction performance, with R² and RMSE of 0.996,3 and 0.001,3, respectively. 【Conclusion】 The results can provide methodological guidance for predicting other nutritional indicators of milk and a reference for studying the quality of other natural emulsions.

Key words: milk; moisture content; dielectric property; particle swarm algorithm; support vector regression; partial least squares regression

中图分类号:

王欢, 张轶腾, 王斌, 周童, 梁清, 张宏. 基于介电特性的生鲜牛奶含水率检测方法[J]. 新疆农业科学, 2025, 62(4): 975-981.

WANG Huan, ZHANG Yiteng, WANG Bin, ZHOU Tong, LIANG Qing, ZHANG Hong. Study on the moisture content detection method of fresh milk based on dielectric property[J]. Xinjiang Agricultural Sciences, 2025, 62(4): 975-981.

图/表 7

图1 生鲜牛奶电参数测量试验台注:1:矢量网络分析仪;2:连接电缆;3:末端开口的同轴探头;4:待测液体

Fig.1 Test bench for Electrical parameters of fresh milk Notes:1,Vector network analyzer;2,Connecting cable;3,End opening coaxial probe;4,Liquid to be tested

表1 生鲜牛乳的主要成分

Tab.1 Main components of the fresh cow milk

参数 Parameter	最小值 Minimum(%)	最大值 Maximum(%)	平均值±标准偏差 Average value± standard deviation(%)
含水率Water content	86.900	92.900	89.900± 2.16
非脂乳固体质量分数Non-fat milk solid mass fraction	4.390	8.100	6.245±1.336
碳水化合物质量分数Carbohydrate mass fraction	2.710	5.000	3.855±0.824
脂肪质量分数Fat mass fraction	1.897	3.500	2.698±0.577
蛋白质质量分数Protein mass fraction	1.789	3.300	2.544±0.544

图2 介电参数随频率的变化规律

Fig.2 Changes of dielectric parameters with frequency

图3 不同变量预测集中含水率的实测值与PLSR模型预测值的比较

Fig.3 Measured values and PLSR of water content of different variables Comparison of the predicted values of the model

表2 不同建模方法对牛奶含水率的预测结果比较

Tab.2 Comparisons of prediction results for different modeling methods

建模方法 Modeling method	建模变量 Mode- ling variable		训练阶段 Training phase				预测阶段 Prediction phase
建模方法 Modeling method	建模变量 Mode- ling variable		R²		RMSE		R²		RMSE
PLSR		ε'		0.994 4		0.001 0		0.867		0.002 4
PLSR		ε″		0.999 3		0.000 4		0.901 5		0.001 8
SVR		ε'		0.993 6		0.001 8		0.978 1		0.004 0
SVR		ε″		0.997 9		0.001 5		0.792 2		0.010 2
PSO- LSSVR		ε'		0.991 7		0.001 7		0.992 0		0.002 6
PSO- LSSVR		ε″		0.995 3		0.001 4		0.996 3		0.001 3

图4 不同变量预测集中含水率的实测值与SVR模型预测值的比较

Fig.4 Comparisons measured values and SVR of water content in different variables Comparison of the predicted values of the model

图5 不同变量预测集中含水率的实测值与PSO-LSSVR模型预测值的比较

Fig.5 Comparisons measured values of predicted content of different variables and PSO-LSSVR Comparison of the predicted values of the model

参考文献 27

[1]	GB 5009.3—2016.食品安全国家标准食品中水分的测定[S].
	GB 5009.3—2016. National Food Safety Standard Determination of Moisture in Food:[S].
[2]	Kasemsumran S, Thanapase W, Kiatsoonthon A. Feasibility of near-infrared spectroscopy to detect and to quantify adulterants in cow milk[J]. Analytical Sciences, 2007, 23(7): 907-910. PMID
[3]	Eltemur D, Robatscher P, Oberhuber M, et al. Applications of solution NMR spectroscopy in quality assessment and authentication of bovine milk[J]. Foods, 2023, 12(17): 3240.
[4]	Mabrook M F, Petty M C. A novel technique for the detection of added water to full fat milk using single frequency admittance measurements[J]. Sensors and Actuators B: Chemical, 2003, 96(1/2): 215-218.
[5]	佀博学, 张养东, 郑楠, 等. 生乳新鲜度评价指标研究进展[J]. 动物营养学报, 2023, 35(6): 3499-3507. DOI
	SI Boxue, ZHANG Yangdong, ZHENG Nan, et al. Research progress on freshness evaluation indices of raw milk[J]. Chinese Journal of Animal Nutrition, 2023, 35(6): 3499-3507. DOI
[6]	曹玉栋, 祁伟彦, 李娴, 等. 苹果无损检测和品质分级技术研究进展及展望[J]. 智慧农业, 2019, 1(3): 29-45.
	CAO Yudong, QI Weiyan, LI Xian, et al. Research progress and prospect on non-destructive detection and quality grading technology of apple[J]. Smart Agriculture, 2019, 1(3): 29-45. DOI
[7]	胡志晨. 基于介电特性的苹果无损分级测量模块的设计与实现[D]. 杨凌: 西北农林科技大学, 2016.
	HU Zhichen. Design and implementation of measurement module for nondestructive classification of apples based on dielectric property[D]. Yangling: Northwest A & F University, 2016.
[8]	Clerjon S, Daudin J D, Damez J L. Water activity and dielectric properties of gels in the frequency range 200 MHz-6 GHz[J]. Food Chemistry, 2003, 82(1): 87-97.
[9]	郭文川, 孔繁荣. 不同类型蛋白质及其添加量对牛乳介电特性的影响[J]. 农业机械学报, 2016, 47(8): 248-254.
	GUO Wenchuan, KONG Fanrong. Influence of different proteins and their additive amounts on permittivities of milk[J]. Transactions of the Chinese Society for Agricultural Machinery, 2016, 47(8): 248-254.
[10]	梁志斌. 生鲜牛乳介电特性与脂肪含量关系的试验研究[D]. 杨凌: 西北农林科技大学, 2014.
	LIANG Zhibin. study of experiment on the relationshiop between dielectric properties and at content of raw milk.[D]. Yangling: Northwest A & F University, 2014.
[11]	Nelson S O, Trabelsi S. Factors influencing the dielectric properties of agricultural and food products[J]. The Journal of Microwave Power and Electromagnetic Energy: a Publication of the International Microwave Power Institute, 2012, 46(2): 93-107.
[12]	Sacilik K, Colak A. Determination of dielectric properties of corn seeds from 1 to 100 MHz[J]. Powder Technology, 2010, 203(2): 365-370.
[13]	张本华, 钱长钱, 焦晋康, 等. 基于介电特性与SPA-SVR算法的水稻含水率检测方法[J]. 农业工程学报, 2019, 35(18): 237-244.
	ZHANG Benhua, QIAN Changqian, JIAO Jinkang, et al. Rice moisture content detection method based on dielectric properties and SPA-SVR algorithm[J]. Transactions of the Chinese Society of Agricultural Engineering, 2019, 35(18): 237-244.
[14]	武新慧, 郭玉明, 孙静鑫, 等. 苹果介电特性与微波干燥含水率相关性研究[J]. 农机化研究, 2018, 40(10): 194-198.
	WU Xinhui, GUO Yuming, SUN Jingxin, et al. Study on coaxial electric properties of apple as a function of microwave drying moisture content[J]. Journal of Agricultural Mechanization Research, 2018, 40(10): 194-198.
[15]	李大伟, 陈超, 兰海鹏, 等. 基于电特性的新疆骏枣含水率的试验研究[J]. 农机化研究, 2016, 38(7): 212-215.
	LI Dawei, CHEN Chao, LAN Haipeng, et al. Experimental research on the electrical characteristics of Xinjiang Junzao moisture[J]. Journal of Agricultural Mechanization Research, 2016, 38(7): 212-215.
[16]	周世平, 张海红, 李海峰, 等. 基于果品介电特性的无损检测技术研究综述[J]. 食品研究与开发, 2015, 36(1): 131-134, 144.
	ZHOU Shiping, ZHANG Haihong, LI Haifeng, et al. Review of fruit nondestructive testing technology research based on the dielectric characteristics[J]. Food Research and Development, 2015, 36(1): 131-134, 144.
[17]	McKeown M S, Trabelsi S, Tollner E W, et al. Dielectric spectroscopy measurements for moisture prediction in Vidalia Onions[J]. Journal of Food Engineering, 2012, 111(3): 505-510.
[18]	郭文川, 林碧莹. 牛奶含水率介电谱结合化学计量学检测方法[J]. 农业机械学报, 2016, 47(9): 249-255.
	GUO Wenchuan, LIN Biying. Detecting moisture content of cow’s milk using dielectric spectra and chemometrics[J]. Transactions of the Chinese Society for Agricultural Machinery, 2016, 47(9): 249-255.
[19]	GB/T 35680—2017.中国国家标准化管理委员会.液体材料微波频段使用开口同轴探头的电磁参数测量方法[S].
	GB/T 35680—2017. Measuring method for electromagnetic parameters of liquid materials at microwave frequencies using an open-ended coaxial probe[S].
[20]	Wold S, Sjöström M, Eriksson L. PLS-regression: a basic tool of chemometrics[J]. Chemometrics and Intelligent Laboratory Systems, 2001, 58(2): 109-130.
[21]	王建国, 张文兴. 支持向量机建模及其智能优化[M]. 北京: 清华大学出版社, 2015: 12-16.
	WANG Jianguo, ZHANG Wenxing. The SVM modeling and its intelligent optimization[M]. Beijing: Tsinghua University Press, 2015: 12-16.
[22]	Chang C C, Lin C J. LIBSVM: a library for support vector machines[J]. ACM Transactions on Intelligent Systems and Technology, 2011, 2(3): 1-27.
[23]	刘昱, 赵国新. 基于粒子群优化算法的无摩擦气缸结构优化[J]. 系统仿真学报, 2018, 30(9): 3564-3570. DOI
	LIU Yu, Zhao Guoxin. Optimization of frictionless cylinder-structure based on PSO[J]. Journal of System Simulation, 2018, 30(9): 3564-3570. DOI
[24]	刘贺, 张弘强, 刘斌. 基于粒子群优化神经网络算法的深基坑变形预测方法[J]. 吉林大学学报(地球科学版), 2014, 44(5): 1609-1614.
	LIU He, ZHANG Hongqiang, LIU Bin. A prediction method for the deformation of deep foundation pit based on the particle swarm optimization neural network[J]. Journal of Jilin University (Earth Science Edition), 2014, 44(5): 1609-1614.
[25]	朱新华, 郭文川. 影响食品射频-微波介电特性的因素及影响机理分析[J]. 食品科学, 2010, 31(17): 410-414. DOI
	ZHU Xinhua, GUO Wenchuan. A review of affecting factors and their mechanisms of the radio frequency-microwave dielectric properties of foods[J]. Food Science, 2010, 31(17): 410-414. DOI
[26]	Zhu X H, Guo W C, Liang Z B. Determination of the fat content in cow’s milk based on dielectric properties[J]. Food and Bioprocess Technology, 2015, 8(7): 1485-1494.
[27]	Zhu X H, Guo W C, Jia Y P, et al. Dielectric properties of raw milk as functions of protein content and temperature[J]. Food and Bioprocess Technology, 2015, 8(3): 670-680.